OpenClaw WeCom Channel

企业微信 (WeCom) Channel 插件 — 让 OpenClaw AI Agent 通过企业微信收发消息。支持消息加解密、Token 自动管理、访问控制策略。

This plugin appears to do what it claims, but take these precautions before installing: 1) Verify the repository origin and review the code (especially token handling and the HTTP server) before cloning. 2) Store WeCom secrets (Agent Secret, Token, EncodingAESKey) only in OpenClaw's config with restrictive file permissions or use environment variables if supported; treat them like sensitive credentials. 3) Expose the callback endpoint through a secure tunnel (Cloudflare Tunnel/ngrok) and ensure TLS and WeCom trusted IP configuration to limit exposure. 4) Restrict plugin usage with the dmPolicy or allowlist if you don't want it open to all users. 5) If you stop trusting the plugin, rotate the WeCom Secret in the enterprise WeCom admin and clear token cache. If you want additional assurance, ask for a signed release or a maintainer identity and check recent commits for unexpected network or filesystem activity.

Type: OpenClaw Skill Name: openclaw-wecom-channel Version: 0.1.0 The OpenClaw WeCom channel plugin is designed to integrate OpenClaw AI agents with the WeCom messaging platform. It handles message sending/receiving, encryption/decryption (using WeCom's WXBizMsgCrypt standard), and access token management. The plugin requires network access and opens an HTTP port for WeCom callbacks, which is explicitly declared and necessary for its functionality. It uses official WeCom API endpoints and handles sensitive credentials (corpId, secret, etc.) as expected for an integration. There is no evidence of intentional malicious behavior such as data exfiltration to unauthorized third parties, unauthorized command execution, or prompt injection attempts against the AI agent. The use of SHA1 for message signing is a requirement of the WeCom API, not a malicious choice by the plugin developer. The code is straightforward and aligns with the stated purpose.